There has never been a greater need for research on new preventative and therapeutic vaccines to prep and treat tuberculosis (TB). BCG, the only currently available vaccine, has variable efficacy and may not be safe in the context of HIV infection. The morality rate from TB is increasing as multi-drug resistant strains of Mycobacterium tuberculosis (Mtb) become more prevalent. The co-prevalence of HIV and TB in the United States and in developing countries will contribute to the acceleration of the TB case rate worldwide. Advances in TB prevention and immunotherapy must be based on identification of the immunodominant epitopes of antigens that promote "protective" immune responses. In this application, I outline a proposal to identify Mtb antigens that are associated with a protective pattern of response in Mtb-infected patients, using two novel experimental approaches. I propose to identify Mtb protein epitopes by extracting "natural peptides" from antigen presenting cells infected with live, virulent Mtb. Additional Mtb epitopes will be identified by screening published Mtb protein sequences for amino acid sequences that conform to established "natural" motifs and/or to a computerized algorithin (AMPHI). These putative epitopes will be-synthesized and tested for immunogenicity in in vitro assays using cells and cell lines derived from Mtb-immune subjects. Peptide epitopes that appear to be immunogenic will then be evaluated in assays designed to compare the immune responses of Mtb-immune and Mtb- susceptible subjects. Peptides that provoke T cell proliferation or CTL responses in cells derived from Mtb-immune subjects, but not in cells derived from Mtb-susceptible subjects, will be considered "protective" antigens, candidates for inclusion in a synthetic multi-subunit TB vaccine. I will also explore the potential use of "protective" antigens in the development of an adoptive immunotherapy technique for the treatment of multi-drug resistant TB. Preliminary investigations described in this application demonstrate a decrease of Mtb-specific T cell proliferation responses to several mycobacterial protein preparations (PPD, lyophilized Mtb) in Mtb- susceptible subjects. I also present results from comparisons of Mtb- specific cytotoxicity in cell lines derived from Mtb-immune and Mtb- susceptible individuals. The isolation of natural peptide epitopes and the identification of peptides by algorithm will be performed in collaboration with Dr. Jay A. Berzofsky of the National Cancer Institute, NIH. Dr. Judy Lieberman of New England Medical Center will collaborate on the development of cell lines for adoptive immunotherapy. The in vitro assays will be performed in the Biohazard level 3 facility at Brown University. The study population will include patients with latent Mtb infection and active TB disease, who will be recruited in three public hospital settings. Indigent individuals, Blacks, and Hispanics will make up a large proportion of the study population.